augustus hints and "No feasible path found in HMM" error message

Discussions about predicting genes with AUGUSTUS. Not covered here: WebAUGUSTUS and BRAKER1

Moderator: bioinf

Post Reply
User avatar
katharina
Site Admin
Posts: 531
Joined: Wed Nov 18, 2015 6:14 pm
Location: Greifswald
Contact:

augustus hints and "No feasible path found in HMM" error message

Post by katharina »

Originally posted in the old forum by Rajat S. Roy on 15.11.2012 - 20:28
I was using augustus with some hints and I keep getting "No feasible path found in HMM"error message for many combination of these hints. It seems like the HMM mechanism is overfitting the hints and some transition/ emission probabilities might be going to zero. Is there a quick way of fixing this?
I was using the following command:

Code: Select all

 augustus --species=generic --hintsfile=augustus_hints.gff --codingseq=on --protein=on chromosomes.fa > aug.gff
the chromosomes are from phaeodactylum tricornutum and the hints come from some genes predicted by augustus in a previous run with no hints.
Thanks
Rajat
User avatar
katharina
Site Admin
Posts: 531
Joined: Wed Nov 18, 2015 6:14 pm
Location: Greifswald
Contact:

Re: augustus hints and "No feasible path found in HMM" error message

Post by katharina »

by Mario on 28.11.2012 - 11:34
Do not use --species=generic for gene predictions!
I understand it is tempting to think that this is a parameter set that works ok for many species without training. But 'generic' is only a template for the meta parameters for training other species.
The actual parameters of generic are (currently) of no use to anyone. We may try to find a consensus parameter set of all eukaryotes, but it is foreseeable that this would not work very well unless supplemented with hints.
Post Reply